DESCRIPTION (taken from abstract): The hypothesis of this proposed study is that diffusion MRI, which measures the apparent diffusion coefficient of tissue water (ADCw) has potential to be a non-invasive marker for effective cancer chemotherapy. A significant amount of pre-clinical evidence has shown that ADCw increases in tumors that are destined to respond to successful chemotherapy, and that it does not change if the therapy is ineffective. It is believed that these local changes in ADCw result from the early stages of apoptosis. The proposal will test the hypothesis in a clinical setting. Breast cancer patients with metastatic liver lesions who are undergoing taxotere therapy will be examined by diffusion MRI prior to, and 5-30 days after, commencement of therapy. Diffusion and T2-weighted images will be obtained. Patients will be followed for 44 days to dteermine the extent of their response by objective and subjective clinical criteria. This trial will demonstrate whether diffusion images predict the final response to therapy.